微波EDA网,见证研发工程师的成长!
首页 > 研发问答 > 测试测量 > 虚拟仪器 > 几个labview中信号处理问题,希望大家来讨论一下

几个labview中信号处理问题,希望大家来讨论一下

时间:12-13 整理:3721RD 点击:
我想用labview做一个带通是2HZ左右的带通滤波器,带阻衰减达40db以上,labview中可以
用parks-mcclellan.vi,fir narrowband filter.vi,等,可是阶数多达1200以上,这样过
渡带太长,需要采集数据太多,就是浪费很多数据!有没有人用过这么高阶的滤波器?如
何把阶数降下来呀?椭圆滤波器过渡带也是很长的!
.250

1200阶??是fs太高了吧
用硬件做一个简单的滤波电路吧
.46

fs我采用1k的,用模拟硬件滤波器精度达不到,而且元件参数极度敏感,我先前仿真过就
是不行呀!用晶体滤波器?
.250

你的采样频率是多少?
我刚才看了一下,如果我设fs=10 Hz, passband freuqencies are 4 and 6 Hz ( then t
he pass band is 2 Hz) and stopband frequencies are 3.5 Hz and 6.5 Hz. the pass
band response -5 dB and stopband response -45dB. Then the Butterworth fiter is
11-order, Elliptic filter is 4-order.
If you need the filter coefficients, I can email you.
.135

我的采样频率蛮大的是1kHz呢,
.250

Compared with your samping rate of 1 KHz, 2Hz band is too small. I suppose you
can use a notch filter than a bandpass filter. It is easy to design this in t
he z plane by setting a pair of pole-zero. For example, the normalized frequen
cy (0, Pi) rad/sec correspond to  your (0, 500)Hz. If your want to filter the
component around 250Hz, then you can use put a zero at  pi close to the origin
and a pole with the same phase but close to the unit circle (inside). Then yo
u are done. If you increase the order to 5, that will be better.
  
.135

Thank you for your valuable advices!
.250

Copyright © 2017-2020 微波EDA网 版权所有

网站地图

Top